Bloomberg the Company & Products

Bloomberg Anywhere Login

Bloomberg

Connecting decision makers to a dynamic network of information, people and ideas, Bloomberg quickly and accurately delivers business and financial information, news and insight around the world.

Company

Financial Products

Enterprise Products

Media

Customer Support

  • Americas

    +1 212 318 2000

  • Europe, Middle East, & Africa

    +44 20 7330 7500

  • Asia Pacific

    +65 6212 1000

Communications

Industry Products

Media Services

Follow Us

Security

Tip for Bad Guys: Burn, Don't Shred


To most people, 10,000 slivers of shredded paper are as good as trash. To three coders in San Francisco, they’re a challenge—especially when the jumbled mass of paper once made up five classified government documents.

The trio were not hackers trying to steal state secrets, but participants in a contest run by the Defense Advanced Research Projects Agency (Darpa), the government group that funds high-tech military research. In October, Darpa offered $50,000 to the first group to piece together the shredded documents or the one that made the most progress by Dec. 4. In previous Darpa tournaments, participants have been asked to build robotic cars or use the Internet to find balloons scattered across the country. The goal of the paper shredder puzzle was to unearth technologies that could be used for national security.

“I figured I know enough really damn good programmers that I could get a few people together and we might be able to win it,” says Otávio Good, the entrepreneur behind the iPhone app Word Lens, which translates foreign-language text as it’s viewed through the phone’s camera. Good and his partners, a software engineer working at Lockheed Martin (LMT) and a mobile app maker, spent 600 hours combined piecing together the five shredded pages. Out of nearly 9,000 teams, theirs—which they called “All Your Shreds Are Belong to U.S.”—was the only one to complete all five puzzles, which they did with two days to spare.

Good’s team relied on a custom-built computer program to analyze the ink, tears, and other markings on the scraps to suggest possible matches, as well as guesswork. “I knew there would be crazy people to try to do it all manually and crazy people to try to automate the whole process,” says Good. “We went with an in-between approach.”

Two weeks in, Good’s team was stumped by the fourth sheet of paper, which was more of a challenge because the words were irregularly spaced and written on unlined paper. “I think everybody hit a wall on puzzle four and didn’t know what to do,” Good says. Then he happened on an article about a little-known government project: The U.S. Secret Service has been working with manufacturers of color laser printers to place tiny, imperceptible yellow dots on printed pages so that the government can track the machine that produced them. Good put the puzzle under a blue-light filter and saw the dots. “It was a breakthrough moment,” he says. “The pattern of dots is practically a map for how everything comes together.”

Good says that the team hasn’t discussed its plans for the $50,000 prize, but they’ve already gotten recognition from peers. “I think we have more programmer street cred,” he says.

The bottom line: Three Bay Area programmers used a mixture of human and computational analysis to win the Darpa Shredder Challenge.

MacMillan is a reporter for Bloomberg News and Bloomberg Businessweek in San Francisco.

LIMITED-TIME OFFER SUBSCRIBE NOW
 
blog comments powered by Disqus